Patient characteristics
From January 2019 to October 2022, 110 patients with stage IV NSCLC were treated with combination therapy including an ICI and platinum doublet chemotherapy. Four patients were excluded from this study because of a lack of information on body weight at 6 and 9 weeks after the initiation of the first course of chemotherapy (Fig. 1). The median follow-up duration was 389 days (95% confidence interval [CI]: 319–470 days).
The median age was 67.5 years, 81 (76.4%) were male, 105 (99.1%) had good PS (ECOG-PS 0–1), and 94 (88.7%) were current or former smokers. The most common histological types were adenocarcinoma (n=82, 77.4%) and squamous cell carcinoma (n=14, 13.2%). Twenty-one (19.8%) patients had more than 5% body weight loss 6 months before the start of first-line chemotherapy. Patient characteristics at the start of the first-line chemotherapy are shown in Table 1.
Table 1. Patient characteristics
|
All patient
(n=106)
|
non-CIA
(n=92)
|
CIA
(n=14)
|
p-value
|
TC
(n=76)
|
non-TC
(n=30)
|
p-value
|
Age (years)
Median (range)
|
67.5
(36–85)
|
67.5
(36–85)
|
66.5
(57–74)
|
0.815
|
68
(36–85)
|
64.5
(50–83)
|
0.277
|
Sex, n (%)
Male
Female
|
81 (76.4)
25 (23.6)
|
69 (75.0)
23 (25.0)
|
12 (85.7)
2 (14.3)
|
0.511
|
57 (75.0)
19 (25.0)
|
24 (80.0)
6 (20.0)
|
0.799
|
ECOG-PS, n (%)
|
0
1
2
|
24 (22.6)
81 (76.4)
1 (0.9)
|
23 (25.0)
68 (73.9)
1 (1.1)
|
1 (7.1)
13 (92.9)
0 (0.0)
|
0.290
|
20 (26.3)
55 (72.4)
1 (1.3)
|
4 (13.3)
26 (86.7)
0 (0.0)
|
0.314
|
Histopathology, n (%)
|
Ad
Sq
Other
|
82 (77.4)
14 (13.2)
10 (9.4)
|
73 (79.4)
11 (12.0)
8 (8.7)
|
9 (64.3)
3 (21.4)
2 (14.3)
|
0.184
|
60 (79.0)
11 (14.5)
5 (6.6)
|
22 (73.3)
3 (10.0)
5 (16.7)
|
0.281
|
PD-L1 status, n (%)
|
≥50%
1-49%
<1%
unknown
|
11 (10.4)
31 (29.2)
59 (55.7)
5 (4.7)
|
10 (10.9)
26 (28.3)
51 (55.4)
5 (5.4)
|
1 (7.1)
5 (35.7)
8 (57.1)
0 (0.0)
|
0.959
|
10 (13.2)
21 (27.6)
40 (52.6)
5 (6.6)
|
1 (3.3)
10 (33.3)
19 (63.3)
0 (0.0)
|
0.236
|
Stage, n (%)
ⅣA
ⅣB
|
51 (48.1)
55 (51.9)
|
45 (48.9)
47 (51.1)
|
6 (42.9)
8 (57.1)
|
0.778
|
40 (52.6)
36 (47.4)
|
11 (36.7)
19 (63.3)
|
0.195
|
Smoking status, n (%)
|
Current/former
Never
|
94 (88.7)
12 (11.3)
|
81 (88.0)
11 (12.0)
|
13 (92.9)
1 (7.1)
|
1.000
|
67 (88.2)
9 (11.8)
|
27 (90.0)
3 (10.0)
|
1.000
|
BMI (kg/m2)
Median (range)
|
21.2
(15.7–35.3)
|
21.7
(15.7–33.9)
|
19.6
(17.4–35.3)
|
0.334
|
21.5
(15.7–35.3)
|
20.6
(16.4–26.3)
|
0.285
|
≥5% Body weight loss in previous 6 months
|
Yes
No
Unknown
|
21 (19.8)
40 (37.7)
45 (42.5)
|
16 (17.4)
35 (38.0)
41 (44.6)
|
5 (35.7)
5 (35.7)
4 (28.6)
|
0.271
|
15 (19.7)
30 (39.5)
31 (40.8)
|
6 (20.0)
10 (33.3)
14 (46.7)
|
0.818
|
Platinum agent
CDDP
CBDCA
|
30 (34.0)
76 (66.0)
|
29 (31.5)
63 (68.5)
|
7 (50.0)
7 (50.0)
|
0.227
|
24 (31.6)
52 (68.4)
|
12 (40.0
18 (60.0)
|
0.496
|
Concomitant anticancer drug, n (%)
|
PEM
nab-PTX
PTX
PTX+Bev
|
81 (76.4)
14 (13.2)
5 (4.7)
6 (5.7)
|
73 (79.4)
11 (12.0)
4 (4.4)
4 (4.4)
|
8 (57.1)
3 (21.4)
1 (7.1)
2 (14.3)
|
0.119
|
58 (76.3)
10 (13.2)
4 (5.3)
4 (5.3)
|
23 (76.7)
4 (13.3)
1 (3.3)
2 (6.7)
|
1.000
|
Concomitant Immune checkpoint inhibitor, n (%)
|
Pembrolizumab
Atezolizumab
Nivo+Ipi
|
89 (84.0)
16 (15.1)
1 (0.9)
|
78 (84.8)
13 (14.1)
1 (1.1)
|
11 (78.6)
3 (21.4)
|
0.518
|
64 (84.2)
11 (14.5)
1 (1.3)
|
25 (80.0)
5 (20.0)
|
0.836
|
Use of Olanzapine on days 1–8 of the first course of treatment, n (%)
|
Yes
No
Preventive use
|
24 (22.6)
82 (77.4)
18 (17.0)
|
18 (19.6)
74 (80.4)
16 (17.4)
|
6 (42.9)
8 (57.1)
2 (14.3)
|
0.081
1.000
|
12 (15.8)
64 (84.2)
6 (7.9)
|
12 (40.0)
18 (60.0)
6 (20.0)
|
0.011
0.579
|
Abbreviation: ECOG-PS, Eastern Cooperative Oncology Group Performance; Ad, adenocarcinoma; Sq, squamous cell carcinoma; PD-L1, programmed cell death ligand 1; BMI, body mass index; CDDP, cisplatin; CBDCA, carboplatin; PEM, pemetrexed; PTX, paclitaxel; Bev, bevacizumab; Nivo, nivolumab; Ipi, ipilimumab.
Frequency of CINV and CIA
Among the 106 patients, 14 (13.2%) and 30 (28.3%) were classified into the CIA and non-TC groups, respectively. Of these, 10 in the CIA group (10/14, 71.4%) were in the non-TC group and 10 in the non-TC group (10/30, 33.3%) were in the CIA group (Pearson's chi-square test, p=0.0001) (Supplementary Table A1). There were no significant between-group differences in clinicopathological characteristics, such as age, sex, ECOG-PS, tumor histology, PD-L1 status, clinical stage, smoking status, baseline BMI, frequency of body weight loss ≥ 5% in the previous 6 months, and drugs used (platinum agents, concomitant anticancer drugs, and ICIs). The use of olanzapine on days 1–8 of the first course of treatment was not significantly different between the CIA and non-CIA groups, but was more frequent in the non-TC than in the TC group. There was no significant group difference in the preventive use of olanzapine.
Dose delivery and therapy duration
There were no significant differences in the type of platinum agent used (CDDP use: CIA, 7 [50.0%] vs. non-CIA, 29 [31.5%]; p=0.23); non-TC, 12 [40.0%] vs. TC, 24 [31.6%]; p=0.50), mean duration of platinum doublet chemotherapy, treatment interval per course, or cycles of platinum doublet chemotherapy.
There was no difference in the median number of cycles for platinum doublet chemotherapy administration: 4 (range: 3–4) in the CIA group, 4 (range: 4–4) in the non-CIA group (p=0.80), 4 (range: 4–4) in the non-TC group, and 4 (range: 3–4) in the TC group (p=0.34). The median number of cycles for maintenance therapy administration was significantly lower in the CIA than in the non-CIA group (5, range: 3–6 vs. 8, range: 5–15; p=0.01), but not in the non-TC and TC groups (4, range: 4–4 vs. 4, range: 3–4; p=0.34).
The relative dose delivery intensity of platinum agents and combination anticancer drugs was 100% in the CIA (interquartile range [IQR]: 100–100) and non-CIA groups (IQR: 87.5–100), and that of concomitant anticancer drugs was 100% in the CIA (IQR: 100–100) and non-CIA groups (IQR: 85.9–100). Notably, both were significantly lower in the CIA group (p<0.01). In addition, the frequency of dose reduction within the first four cycles for platinum agents and concomitant anticancer drugs was significantly higher in the CIA group than in the non-CIA group (platinum agent: 2.2% vs. 28.6%, p<0.01; concomitant anticancer drugs: 4.4% vs. 28.6%, p=0.01). No significant differences in the frequency of dose reduction were observed between the TC and non-TC groups (platinum agent: 4.0% vs. 10.0%, p=0.35. concomitant anticancer drugs: 5.3% vs. 13.3%, p=0.22). There was no significant between-group difference in the frequency of dose delays of ≥1 week or treatment cessation (Table 2).
Table 2. Dose intensity by CIA/TC group
|
non-CIA
(n=92)
|
CIA
(n=14)
|
p-value
|
TC
(n=76)
|
non-TC
(n=30)
|
p-value
|
Concomitant platinum-agent, n (%)
CDDP
CBDCA
|
29 (31.5)
63 (68.5)
|
7 (50.0)
7 (50.0)
|
0.23
|
24 (31.6)
52 (68.4)
|
12 (40.0)
18 (60.0)
|
0.50
|
Mean duration of platinum-doublet therapy [weeks]
Median (IQR)
|
9.7
(9–11.4)
|
9.9
(8.9–12.0)
|
0.11
|
9.1
(9–11)
|
10.1
(9.2–12)
|
0.07
|
Treatment interval / course (days)
Median (range)
|
23.3
(21–27.3)
|
23.0
(21.3–28.1)
|
0.90
|
22.8
(21–27.3)
|
23.5
(21.3–28)
|
0.42
|
Cycles for platinum-doublet therapy
Median (IQR)
|
4
(4–4)
|
4
(3–4)
|
0.80
|
4
(4–4)
|
4
(4–4)
|
0.34
|
Cycles for maintenance therapy
Median (IQR)
|
8
(5–15)
|
5
(3–6)
|
0.01
|
8
(4–15)
|
6
(5–10)
|
0.25
|
Platinum-agent
|
|
|
|
|
|
|
Relative dose intensity (%)
Median (IQR)
|
100
(100–100)
|
100
(87.5–100)
|
<0.01
|
100
(100–100)
|
100
(100–100)
|
0.22
|
Dose reductiona [within 4 cycles], n (%)
|
2 (2.2)
|
4 (28.6)
|
<0.01
|
3 (4.0)
|
3 (10.0)
|
0.35
|
Combination anticancer drug
|
|
|
|
|
|
|
Relative dose intensity (%)
Median (IQR)
|
100
(100–100)
|
100
(85.9–100)
|
<0.01
|
100
(100–100)
|
100
(100–100)
|
0.14
|
Dose reductionb [within 4 cycles], n (%)
|
4 (4.4)
|
4 (28.6)
|
0.01
|
4 (5.3)
|
4 (13.3)
|
0.22
|
Dose delays ≥1 week / cycle [within 4 cycles], n (%)
|
47 (51.1)
|
5 (35.7)
|
0.39
|
36 (47.4)
|
16 (53.3)
|
0.67
|
Treatment cessation [not PD, within 4 cycles], n (%)
|
12 (13.0)
|
3 (21.4)
|
0.42
|
11 (14.5)
|
4 (13.3)
|
1.00
|
aReasons for dose reductions: Grade 3 anorexia (3), Grade 3 febrile neutropenia (2), and Grade 2 rash (1).
bReasons for dose reductions: Grade 3 anorexia (3), Grade 3 febrile neutropenia (2), Grade 2 rash (2), and Grade 2 peripheral sensory neuropathy.
Impact of CIA and CINV on body weight and BMI
The change in weight from baseline was significantly different between the CIA and non-CIA groups 6 and 9 weeks after treatment initiation (6 weeks: -4.28% vs. 0.36%, p=0.012; 9 weeks: -2.90% vs. 2.20%, p=0.007) (Fig. 2a), but not between the non-TC and TC groups (6 weeks: 0.39% vs. 0.34%, p=0.988; 9 weeks: 2.66 vs. 1.71, p=0.492) (Fig. 2b).
The change in BMI from baseline was also significantly different between the CIA and non-CIA groups 6 and 9 weeks after treatment initiation (6 weeks: -0.79 vs. 0.07, p=0.011; -0.56 vs. 0.45; p=0.011) (Fig. 2c), but not between the non-TC and TC groups (6 weeks: 0.07 vs. 0.07, p=0.994; 9 weeks: 0.47 vs. 0.40, p=0.627) (Fig. 2d, Table 3).
Table 3. Body weight and BMI
|
non-CIA
(n=92)
|
CIA
(n=14)
|
p-value
|
TC
(n=76)
|
non-TC
(n=30)
|
p-value
|
Change in weight from baseline (%)
|
6 weeks after treatment initiation
(IQR)
(missing value=2)
|
0.36
(-1.75–2.73)
|
-4.28
(-8.89–0.47)
|
0.012
|
0.34
(-2.56–2.57)
|
0.39
(-3.02–2.39)
|
0.988
|
9 weeks after treatment initiation
(IQR)
(missing value=4)
|
2.20
(-1.14–4.44)
|
-2.90
(-11.0–3.29)
|
0.007
|
1.71
(-2.30–3.64)
|
2.66
(-1.85–4.72)
|
0.492
|
Change in BMI from baseline (kg/m2)
|
6 weeks after treatment initiation
(IQR)
|
0.07
(-0.40–0.63)
|
-0.79
(-1.93–0.09)
|
0.011
|
0.07
(-0.57–0.55)
|
0.07
(-0.61–0.47)
|
0.994
|
9 weeks after treatment initiation
(IQR)
|
0.45
(-0.23–0.96)
|
-0.56
(-2.27–0.64)
|
0.007
|
0.40
(-0.55–0.86)
|
0.47
(-0.38–0.95)
|
0.627
|
Abbreviations: BMI, body mass index; CIA, chemotherapy-induced anorexia; IQR, interquartile range; TC, total control.
Efficacy
At the time of data cutoff (October 20, 2023), the median follow-up duration from the start of first-line chemotherapy for censored cases was 389 days (95% CI: 319–470 days). Of the 106 included patients, 36 (34.0%) had died at the time of the data cutoff. For the entire cohort, the median PFS was 236.5 days (95% CI: 176–290 days). Although the non-CIA group showed better PFS than the CIA group (median PFS: 258.5 days vs. 123.5 days, HR: 0.57, 95% CI: 0.31–1.03, p=0.06; Fig. 3a), the difference was not statistically significant. There was no significant difference in PFS between the TC and non-TC groups (median PFS: 236.5 days vs. 237 days; HR, 0.70; 95% CI: 0.45–1.10, p=0.12; Fig. 3b).
OS was also better in the non-CIA group than in the CIA group (median OS: 640 vs. 394.5 days, HR: 0.67, 95% CI: 0.35–1.29, p=0.23; Fig. 3c), however, the difference was not statistically significant. In addition, there was no significant difference in OS between the TC and non-TC groups (median OS, 458 vs. 610 days; HR, 0.97; 95% CI: 0.58–1.63, p=0.91; Fig. 3d).
The ORR and PFS stratified according to CIA/non-CIA and TC/non-TC groups are shown in Supplementary Table A1. Among patients with TC (n=76), the non-CIA group had a higher ORR (48.6% vs. 25.0, p=0.617), although the difference was not statistically significant, and longer PFS (262 days vs. 91 days, p<0.0001) than the CIA group. Among patients with non-TC (n=30), both groups had a similar ORR (55.0% vs. 60.0%, p=1.000) and PFS (233 days vs. 256.5 days, p=0.989).